(18g) Evaluating Candidates for New Zeolites with Machine Learning
AIChE Annual Meeting
2020
2020 Virtual AIChE Annual Meeting
Materials Engineering and Sciences Division
Experimental and Computational Approaches to Accelerate and Discover Inorganic Materials
Monday, November 16, 2020 - 9:15am to 9:30am
Synthesizing new zeolites, which are useful for applications like gas separation and catalysis, with specific properties is an ongoing challenge in the zeolite community. While many zeolites have been computationally designed, relatively few of these new zeolites have been experimentally synthesized. Therefore, the goal of this work is to identify hypothetical zeolites that show the most promise as candidates for experimental synthesis. We begin by constructing a new kind of “atlas†of local atomic environments [1] comprising several thousand all-silica zeolites from the Deem “SLC PCOD†database [2] using the Smooth Overlap of Atomic Positions (SOAP) representation [3] as a structural descriptor, Kernel Ridge Regression (KRR) [4] for machine learning of properties, and Kernel Principal Component Analysis (KPCA) [5] for visualizing structural diversity. We evaluate the utility of this atlas by examining correlations between the locations of the atomic environments in the atlas and their energy and volume contributions to their parent frameworks. We then extend this analysis by comparing the hypothetical zeolites of the Deem database to the realized zeolites in the International Zeolite Association (IZA) Database of Zeolite Structures [6]. In particular, we aim to determine whether the structural diversity in the Deem database comprises the majority of the structural characteristics found in the IZA database.
[1] B. A. Helfrecht, R. Semino, G. Pireddu, S. M. Auerbach, M. Ceriotti, J. Chem. Phys. 151, 154112 (2019)
[2] R. Pophale, P. A. Cheeseman, M. W. Deem, Phys. Chem. Chem. Phys 13, 12407-12412 (2011)
[3] A. P. Bartók, R. Kondor, G. Csányi, Phys. Rev. B. 87, 184115 (2013)
[4] A. J. Smola, B. Schölkopf, ICML ‘00 Proc. 17th Intl. Conf. On Machine Learning, 911-918 (2000)
[5] B. Schölkopf, A. Smola, K.-R. Müller, Neural Computation 10, 1299-1319 (1998)
[6] Ch. Baerlocher and L. B. McCusker, Database of Zeolite Structures, http://www.iza-structure.org/databases